温度植被干旱指数(TVDI)与多因子关系研究

利用EOS/MODIS数据,采用归一化植被指数(NDVI)和地表温度(LST)构建NDVI-Ts特征空间,依据该特征空间计算温度植被干旱指数(TVDI)。通过对2009年6期冬小麦数据比较和多因子相关分析,认为TVDI与LST为极显著正相关关系,与NDVI相关性次之。气象因子中降水量(JSL)因子与TVDI相关性较为显著,在作物生长中后期,降水距平(JSJP)因子影响日趋明显;其它气象因子及海拔、灌溉与否等因子作用不显著,在进行大尺度干旱监测时基本可以忽略。如将TVDI与影响较为明显的因子组合建立新的指标或许是一个更好的监测方法。

Study on the Correlation between Temperature Vegetation Dryness Index (TVDI) and Various Factors

To improve the accuracy used in drought monitoring by remote sensing,the monitoring model has been deeply analyzed.Vegetation index (NDVI) and land surface temperature (LST) are combined to construct NDVI-Ts space for drought monitoring using the TVDI (Temperature Vegetation Dryness Index) to indicate the severity of drought,which are based on EOS/MODIS data.In the major winter-wheat-growing areas,six periods' data during different phases of winter wheat,including remote sensing data,meteorological data and so on,have been used for comparison and correlation analysis.The methods include simple correlation analysis,biased correlation analysis and path analysis.Conclusion has been drawn that LST is significantly correlated with TVDI,NDVI is inferior to LST,which is instable.Among meteorological factors,precipitation and TVDI has a significant correlation and it has gradually affected TVDI during the latter phases of winter wheat.However other factors (including sunlight,temperature anomaly,mean temperature,height,irrigation measures,etc.) are of low relativity to TVDI,which may be ignored in drought monitoring on a large scale.In addition,the correlation coefficient between TVDI and soil moisture in 10 cm depth is higher than that of 20 cm depth.Therefore if soil moisture is selected to act as an inversion parameter,the value in 10 cm depth should be paid more close attention to.If TVDI is used to be combined with other factor,which influences remarkably,such as precipitation,maybe it is a better method for drought monitoring.